Re: OPSAREA strategy

Dave CROCKER wrote:
> 
> 
> David Harrington wrote:
>> The IETF is now faced with a problem. We now have multiple protocols,
>> and most defined their own information models, data modeling language
>> and data models, and we are not sure how to correlate information
>> available from different protocols. And while some in the IETF think
>> that is an important problem to resolve, and standardization is the
>> obvious answer, real world operators have been using multiple
>> protocols successfully without vendor-neutral standards for
>> cross-protocol correlation. 
> 
> 
> Since the core reason for trying to have a single standard for a single
> function is to avoid the NxM combinatorial explosion problem, the
> obvious question is why operators haven't been enthusiastic about
> simplifying their data modeling world?
> 

job security.

The current approach (mostly CLI) is ad-hoc and not very
automation-friendly.  Operations automation is controlled
by Unix scripting wizards (and/or kiddies), and it's
probably in-house, poorly documented, and poorly maintained.
It is certainly proprietary, and probably too expensive.
That's the POV of the managers.  To the actual operators,
the current situation is just fine.

The costs related to moving away from this sort
of system (to anything else) are huge,
especially since the very people who need
to execute the transition do not think there is
any need for change.

The costs related to being locked in to such a system
are not as obvious, but people are starting to figure it out.

> One possibility is that the real-world locus of control rests with the
> vendors of the different models.  Hence, the operators really cannot
> influence things. To some extent, getting a critical mass of operators
> to band together might be able to change the dynamic.
> 
> Another is that they have enough history making parallel specifications
> work that they are used to it.  This is where an extremely strong and
> pragmatic value proposition might make a difference.
> 

Free software certainly influences what operators
use to do their job, and that aspect is very positive
and getting better all the time.  Operators use tools,
not standards.

> d/
> 

Andy